MRI-driven turbulent transport: the role of dissipation, channel modes and their parasites

TL;DR

This study investigates how dissipation affects MRI-driven turbulent transport, revealing different regimes depending on the magnetic Prandtl number and showing that transport is not primarily driven by linear channel modes.

Contribution

It provides a refined analysis of dissipation's role in MRI turbulence, highlighting the regimes of transport dependence and the limited influence of linear and parasitic modes.

Findings

Transport depends on Reynolds number for Pm ≤ 1

Transport depends mainly on Pm for Pm > 1 at high Reynolds numbers

Transport is not primarily driven by axisymmetric channel modes

Abstract

In the recent years, MRI-driven turbulent transport has been found to depend in a significant way on fluid viscosity $\nu$ and resistivity $\eta$ through the magnetic Prandtl number $Pm=\nu/\eta$. In particular, the transport decreases with decreasing $Pm$; if persistent at very large Reynolds numbers, this trend may lead to question the role of MRI-turbulence in YSO disks, whose Prandtl number is usually very small. In this context, the principle objective of the present investigation is to characterize in a refined way the role of dissipation. Another objective is to characterize the effect of linear (channel modes) and quasi-linear (parasitic modes) physics in the behavior of the transport. These objectives are addressed with the help of a number of incompressible numerical simulations. The horizontal extent of the box size has been increased in order to capture all relevant (fastest growing) linear and secondary parasitic unstable modes. The major results are the following: i- The increased accuracy in the computation of transport averages shows that the dependence of transport on physical dissipation exhibits two different regimes: for $Pm \lesssim 1$, the transport has a power-law dependence on the magnetic Reynolds number rather than on the Prandtl number; for $Pm > 1$, the data are consistent with a primary dependence on $Pm$ for large enough ($\sim 10^3$) Reynolds numbers. ii- The transport-dissipation correlation is not clearly or simply related to variations of the linear modes growth rates. iii- The existence of the transport-dissipation correlation depends neither on the number of linear modes captured in the simulations, nor on the effect of the parasitic modes on the saturation of the linear modes growth. iv- The transport is usually not dominated by axisymmetric (channel) modes